nanomaterials
Review
Nanoparticles Engineering by Pulsed Laser Ablation
in Liquids: Concepts and Applications
Enza Fazio1,* , Bilal Gökce2 , Alessandro De Giacomo3,4 , Moreno Meneghetti5 , Giuseppe Compagnini6, Matteo Tommasini7 , Friedrich Waag2, Andrea Lucotti7, Chiara Giuseppina Zanchi7 , Paolo Maria Ossi8, Marcella Dell’Aglio4 , Luisa D’Urso6,
Marcello Condorelli6, Vittorio Scardaci6 , Francesca Biscaglia5, Lucio Litti5 , Marina Gobbo5, Giovanni Gallo1, Marco Santoro9, Sebastiano Trusso10 and Fortunato Neri1
1 Department of Mathematical and Computational Sciences, Physics and Earth Physics, University of Messina, Viale F. Stagno D’Alcontres 31, I-98166 Messina, Italy; ggallo@unime.it (G.G.); fneri@unime.it (F.N.) 2 Department of Technical Chemistry I and Center for Nanointegration Duisburg-Essen, University of
Duisburg-Essen, Universitätsstrasse 7, 45141 Essen, Germany; bilal.goekce@uni-due.de (B.G.); friedrich.waag@uni-due.de (F.W.)
3 Department of Chemistry, University of Bari, Via Orabona 4, 70126 Bari, Italy; alessandro.degiacomo@uniba.it
4 CNR-NANOTEC, c/o Department of Chemistry, University of Bari, Via Orabona 4, 70126 Bari, Italy; marcella.dellaglio@imip.cnr.it
5 Department of Chemical Sciences, University of Padova, Via Marzolo 1, 35131 Padova, Italy; moreno.meneghetti@unipd.it (M.M.); francesca.biscaglia@unipd.it (F.B.); lucio.litti@unipd.it (L.L.); marina.gobbo@unipd.it (M.G.)
6 Department of Chemical Sciences, University of Catania, V.le A. Doria 6, 95125 Catania, Italy; gcompagnini@unict.it (G.C.); ldurso@unict.it (L.D.); marcello.condorelli@unict.it (M.C.); vittorio.scardaci@unict.it (V.S.)
7 Department of Chemistry, Materials, Chemical Engineering, Politecnico di Milano, Piazza Leonardo da Vinci 32, I-20133 Milano, Italy; matteo.tommasini@polimi.it (M.T.); andrea.lucotti@polimi.it (A.L.); chiara.zanchi@tiscali.it (C.G.Z.)
8 Department of Energy & Center for NanoEngineered Materials and Surfaces—NEMAS, Politecnico di Milano, Piazza Leonardo da Vinci 32, I-20133 Milano, Italy; paolo.ossi@polimi.it 9 STMicroelectronics S.R.L., Stradale Primosole 37, 95121 Catania, Italy; marco.santoro@st.com 10 CNR-IPCF Istituto per i Processi Chimico-Fisici, 98053 Messina, Italy; trusso@ipcf.cnr.it
* Correspondence: enfazio@unime.it; Tel.:+39-090-676-5394
Received: 28 September 2020; Accepted: 16 November 2020; Published: 23 November 2020
Abstract:Laser synthesis emerges as a suitable technique to produce ligand-free nanoparticles, alloys and functionalized nanomaterials for catalysis, imaging, biomedicine, energy and environmental applications. In the last decade, laser ablation and nanoparticle generation in liquids has proven to be a unique and efficient technique to generate, excite, fragment and conjugate a large variety of nanostructures in a scalable and clean way. In this work, we give an overview on the fundamentals of pulsed laser synthesis of nanocolloids and new information about its scalability towards selected applications. Biomedicine, catalysis and sensing are the application areas mainly discussed in this review, highlighting advantages of laser-synthesized nanoparticles for these types of applications and, once partially resolved, the limitations to the technique for large-scale applications.
Keywords: colloids; laser synthesis; plasmonics; sensing; biomedicine; catalysis; pollutants
1. Introduction
The intrinsic properties of nanoparticles (NPs), possibly combined with other materials, disclose many applications where one can achieve miniaturization (e.g., of electronic equipment), weight reduction (as a result of an increased material efficiency) and/or improved functionalities of materials (e.g., higher durability, conductivity, thermal stability, solubility, reduced friction, selective molecular detection) [1,2]. The remarkable size-tunable properties of nanomaterials produced by laser–matter interaction (e.g., size distribution, agglomeration state/dispersion, crystal structure, surface area and porosity, surface charge, shape/morphology, dissolution/solubility) make them a hot research topic in material science, with far-reaching applications, ranging from quantum computers to cures for cancer [3,4].
Bottom-up and top-down procedures are the two approaches used for the synthesis of nanomaterials [5]. The first include the miniaturization of materials and components (up to the atomic level) with subsequent self-assembling that leads to the formation of nanostructures. Typical examples are the formation of quantum dots during epitaxial growth, or the production of NPs as colloidal dispersions by chemical approaches [6]. Instead, top-down approaches use macroscopic starting structures that are externally controlled by the processing down to nanostructures. Typical examples are etching (controlled by masks), ball milling, metal-organic vapor phase epitaxy and the application of severe plastic deformations [7].
There are many literature papers on laser interaction with hard, soft and smart materials, targeting future applications in the fields of energy production (nano-energy) and biomedicine [8], as well as recent progress in the understanding of the fundamental mechanisms involved in laser processing [9]. Such laser techniques are interesting in many regards, as they enable the processing of photovoltaic cells [10], thermoelectric materials and devices [11], micro and nanosystems for energy storage and conversion [12,13], biodegradable and biocompatible NPs for food packaging [14] as well as vectors for drug and gene delivery [15,16]. The interest towards pure and electrostatically stabilized nanocolloids is well recognized [17–21]. Although nanotechnologies and nanomaterials have much potential to introduce innovative products and production processes in the food industry and in the biomedical field, they are facing major challenges in being cost-effective in the production of e.g., edible and nontoxic nano-delivery systems, or effective formulations that are safe for human consumption and drug treatments [22].
Laser synthesis of colloids, powered by robust, high-power lasers, appears to be a key enabling process that is chemically clean and environmentally friendly, and appealing [23,24] for industrial manufacturing of functional nanomaterials while being useful in many different areas, such as: hydrogen generation [16], hydrogen storage [25], heterogeneous catalysis using colloidal high-entropy alloy NPs [26,27], anticancer [28] and antimicrobial [29] research, drug monitoring [30], additive manufacturing applications [31,32], and nonlinear nanophotonics [33]. In addition, NPs prepared by the Laser Ablation in Liquids (LAL) have been recently used for various and unique applications like friction reduction [34], solar nanofluids [35,36], optical limiting devices [37,38] and so on.
The generation of NPs with LAL still has some challenging aspects, such as the fabrication of NPs with a specific size and shape, the reduction of polydispersity and the increase of productivity [39]. Despite some unsolved problems of a physical, chemical and technical nature, several strategies have been proposed to overcome the above issues, including the selection of the appropriate liquid or stabilizing agent, the optimization of the focusing conditions and liquid levels, as well as the adoption of scanning and fluid dynamics strategies by different liquid handling configurations [40–44] and the postirradiation of colloids [45].
To explore the wide range of opportunities that LAL brings, the unique properties of femtosecond lasers is a further hot research topic, not only to increase NP production, but also to generate structural modifications and new material phases [46,47]. As reported by Streubel et al. [48], low productivity shortcomings in LAL have been almost entirely remedied, reaching NP production rates of several grams per hour. In addition to LAL, Laser Melting in Liquids (LML), Laser Fragmentation in Liquids
Nanomaterials 2020, 10, 2317 3 of 50
(LFL) as well as pulsed Laser Photoreduction/oxidation in Liquids (LPL), offer alternative routes to obtain colloids with controlled NP sizes. Nevertheless, an industrial series product, or manufacturing process based on laser-synthesized particles is yet to come. Together with the development of new strategies to increase NP productivity, future efforts should be directed to further improve the surface properties of the produced NPs, since each application requires a different surface/interface chemistry. This would ensure reaching an advanced stage in those applications where the surface processes determine the performance of the final devices [49].
In this review, we first give an overview of the fundamentals of pulsed laser ablation synthesis. Moreover, we aim to display a combined picture of the several processes implied by LAL [50,51], remarking, with respect to previous reviews [23,39,52–54], the high complexity and multiple time-scale transient physics and chemistry typical of LAL. We point at this target by taking into account different experimental conditions. The fundamental mechanisms of LAL are analyzed in depth, to stimulate concepts that are suitable to develop high performance devices based on the peculiar properties that the nanomaterials produced by LAL can offer. In the second part of this review, we outline selected applications of LAL in biomedicine, catalysis and sensing. We outline the advantages of laser synthesized NPs for these applications (see Figure1) and the issues that have to be solved to approach the high degree of process control that is required in industrial applications.
Nanomaterials 2020, 10, x 3 of 50
offer alternative routes to obtain colloids with controlled NP sizes. Nevertheless, an industrial series product, or manufacturing process based on laser-synthesized particles is yet to come. Together with the development of new strategies to increase NP productivity, future efforts should be directed to further improve the surface properties of the produced NPs, since each application requires a different surface/interface chemistry. This would ensure reaching an advanced stage in those applications where the surface processes determine the performance of the final devices [49].
In this review, we first give an overview of the fundamentals of pulsed laser ablation synthesis. Moreover, we aim to display a combined picture of the several processes implied by LAL [50,51], remarking, with respect to previous reviews [23,39,52–54], the high complexity and multiple time-scale transient physics and chemistry typical of LAL. We point at this target by taking into account different experimental conditions. The fundamental mechanisms of LAL are analyzed in depth, to stimulate concepts that are suitable to develop high performance devices based on the peculiar properties that the nanomaterials produced by LAL can offer. In the second part of this review, we outline selected applications of LAL in biomedicine, catalysis and sensing. We outline the advantages of laser synthesized NPs for these applications (see Figure 1) and the issues that have to be solved to approach the high degree of process control that is required in industrial applications.
Throughout the review we pointed at two main objectives. First, to offer an analytical overview that is updated and complete on the principles of LAL method, as well as on the role that is played on the synthesis of NPs by process parameters, both laser and geometry related, and by the effects specifically depending on the liquid. Second, we describe the applications we presently believe are most promising in the field: these include plasmonic-related sensing and nanomedicine, and catalysis.
Figure 1. Schematic showing the outline of the review. From the fundamentals and up-scaling, in the
center, arrows point outwards to the four application areas.
2. LAL Synthesis Methods: Principle, Process Parameters and Liquids Effects on Nanoparticle Formation
2.1. Mechanism in ns-Pulsed Laser Ablation in Liquid for the Production of Metallic Nanopar ticles
Laser Ablation in Liquid (LAL) for the production of nanostructures is based on the ejection of material by a laser pulse irradiating a solid target immersed in liquid. The laser matter interaction
Figure 1. Schematic showing the outline of the review. From the fundamentals and up-scaling, in the center, arrows point outwards to the four application areas.
Throughout the review we pointed at two main objectives. First, to offer an analytical overview that is updated and complete on the principles of LAL method, as well as on the role that is played on the synthesis of NPs by process parameters, both laser and geometry related, and by the effects specifically depending on the liquid. Second, we describe the applications we presently believe are most promising in the field: these include plasmonic-related sensing and nanomedicine, and catalysis.
2. LAL Synthesis Methods: Principle, Process Parameters and Liquids Effects on Nanoparticle Formation
2.1. Mechanism in ns-Pulsed Laser Ablation in Liquid for the Production of Metallic Nanoparticles
Laser Ablation in Liquid (LAL) for the production of nanostructures is based on the ejection of material by a laser pulse irradiating a solid target immersed in liquid. The laser matter interaction and the consequent ablation are strongly dependent on the irradiance and the duration of the pulse, on the background liquid, on the sample geometry and morphology as well as on the focusing condition. A full description of the LAL, at different experimental conditions, is beyond the scope of this paper and the interested reader can refer to the following review papers: [39,52,55], and references therein. Here, in order to offer a general viewpoint on the complexity of the sequence of the processes involved in LAL, ns-LAL, for the production of metal NPs in water, will be discussed.
As it has been clearly established, ns-PLAL is based on a sequence of different processes: laser ablation and plasma induction, energy exchange from plasma to the liquid and consequent generation of the cavitation bubble and release of particles from the bubble to the solution. In Figure2 the different LAL steps are summarized.
Nanomaterials 2020, 10, x 4 of 50
and the consequent ablation are strongly dependent on the irradiance and the duration of the pulse, on the background liquid, on the sample geometry and morphology as well as on the focusing condition. A full description of the LAL, at different experimental conditions, is beyond the scope of this paper and the interested reader can refer to the following review papers: [39,52,55], and references therein. Here, in order to offer a general viewpoint on the complexity of the sequence of the processes involved in LAL, ns-LAL, for the production of metal NPs in water, will be discussed.
As it has been clearly established, ns-PLAL is based on a sequence of different processes: laser ablation and plasma induction, energy exchange from plasma to the liquid and consequent generation of the cavitation bubble and release of particles from the bubble to the solution. In Figure 2 the different LAL steps are summarized.
Figure 2. Different steps of Laser Ablation in Liquid (LAL). The picture of laser-matter interaction
refers to a ns-laser focused on Pt target in water. The plasma and bubble images and volumes as functions of the delay time since the laser pulse have been acquired with optical emission imaging and shadowgraphy, respectively (Al target in MilliQ water, Elaser = 270 mJ, laser crater size = 1.6 ± 0.2
mm, λlaser = 532 nm, laser pulse duration = 6 ns, laser frequency = 10 Hz). The nanoparticles (NPs)
visible in cuvette and micrograph image are Au NPs.
Laser-matter interaction: The basic mechanisms inducing the laser ablation in liquid, as well as their dependence on laser pulse properties, do not differ notably, with respect of laser ablation in gaseous environment [54]. In the case of ns-laser ablation, just a portion of the laser pulse reaches directly the target surface, while most of the laser pulse is spent in electron heating by inverse bremsstrahlung. This implies that the ablated material is converted to a plasma phase during the laser pulse irradiation. Differently to what can be observed in a gas background environment, as a consequence of the water incompressibility, the ablated material is strongly confined and in turn reaches high density. This effect decreases the penetration of the laser through the plasma during the initial stage of expansion, inducing the propagation of gradient of temperature in the ejected material, i.e the plasma. For what concerns the type of target used, it should be considered that morphological, electrical and optical properties of the target affect the initial interaction of the material with the laser pulse. For example, a metallic target allows the use of lower ablation irradiance with respect to the semiconductor, if we neglect the difference in reflectivity. In any case if an irradiance well above the breakdown threshold is used, the ablation can be assumed congruent, and similar plasma behavior can be observed.
The mechanisms concerning the evolution of the ejected material after laser ablation are strongly correlated to the duration of the laser pulse. Ideally, with a laser pulse with a shorter duration than a few tens of fs, the effect of the interaction of the laser pulse with the ejected material can be neglected and the evolution of the ablated material can be investigated with atomistic models [56]. On the contrary, when the laser pulse is long enough (as in the case of ns) to effectively interact with the ablated material, inducing further ionization and electron heating, the role of the plasma becomes
Figure 2. Different steps of Laser Ablation in Liquid (LAL). The picture of laser-matter interaction
refers to a ns-laser focused on Pt target in water. The plasma and bubble images and volumes as functions of the delay time since the laser pulse have been acquired with optical emission imaging and shadowgraphy, respectively (Al target in MilliQ water, Elaser= 270 mJ, laser crater size = 1.6 ± 0.2 mm, λlaser= 532 nm, laser pulse duration = 6 ns, laser frequency = 10 Hz). The nanoparticles (NPs) visible in cuvette and micrograph image are Au NPs.
Laser-matter interaction: The basic mechanisms inducing the laser ablation in liquid, as well as their dependence on laser pulse properties, do not differ notably, with respect of laser ablation in gaseous environment [54]. In the case of ns-laser ablation, just a portion of the laser pulse reaches directly the target surface, while most of the laser pulse is spent in electron heating by inverse bremsstrahlung. This implies that the ablated material is converted to a plasma phase during the laser pulse irradiation. Differently to what can be observed in a gas background environment, as a consequence of the water incompressibility, the ablated material is strongly confined and in turn reaches high density. This effect decreases the penetration of the laser through the plasma during the initial stage of expansion, inducing the propagation of gradient of temperature in the ejected material, i.e the plasma. For what concerns the type of target used, it should be considered that morphological, electrical and optical properties of the target affect the initial interaction of the material with the laser pulse. For example, a metallic target allows the use of lower ablation irradiance with respect to
Nanomaterials 2020, 10, 2317 5 of 50
the semiconductor, if we neglect the difference in reflectivity. In any case if an irradiance well above the breakdown threshold is used, the ablation can be assumed congruent, and similar plasma behavior can be observed.
The mechanisms concerning the evolution of the ejected material after laser ablation are strongly correlated to the duration of the laser pulse. Ideally, with a laser pulse with a shorter duration than a few tens of fs, the effect of the interaction of the laser pulse with the ejected material can be neglected and the evolution of the ablated material can be investigated with atomistic models [56]. On the contrary, when the laser pulse is long enough (as in the case of ns) to effectively interact with the ablated material, inducing further ionization and electron heating, the role of the plasma becomes dominant and kinetics models coupled with fluidynamics are required for a full description of the phenomenon [57]. In this review, when describing fundamental aspects, we will refer to ns-ablation, while a brief discussion on ultrashort ablation is reported in Section2.2.
Plasma phase: The plasma phase plays an important role during ns-LAL because the plasma is the source of material that allows the formation of NPs. Plasma phase during ns-LAL can be investigated with Optical Emission Spectroscopy (OES) in order to estimate the plasma parameters such as electron number density, Ne, atomic number density, Na, and excitation temperature, T, as well as the features
of plasma dynamics after laser ablation. The emission spectrum of the Laser Induced Plasma (LIP) under water is characterized mainly by continuum radiation with possibly some resonance peaks of the element constituting the target [40]. The absence of the typical atomic spectral lines of the LIP expanding in a gaseous environment is due to the high density reached by the ablated matter under liquid confinement as discussed in [40–42]. In this case, the emission spectrum has a Planck-like shape that has been ascribed to the radiative recombination and that in turn is proportional to the blackbody law, although the spectrum is not related to any blackbody system. To better understand this crucial point, let us consider blackbody (BB) power density in the Wiens’ approximation at hν/kT >> 1, which is valid for the range 400–500 nm and T ~ 6000 K:
Bv(v, T) =2hv 3 c2 1 exp(kThv)− 1 ≈ 2hv 3 c2 exp(− hv kT)∝v 3exp(−hv kT) (1)
where h is the Planck constant,ν is the frequency, c is the light speed, k is the Boltzmann constant and T is the temperature.
On the other hand, the radiative recombination power density is given by the following equation [6]:
ρR p(v, T) =2 r 2 π h4 m3/2e c2 2p2( 1 kT) 3/2 exp(−hv −(Ei −Ep) kT )σion(p, v)v 3N2 e (2)
where me is the electron mass, Ei and Ep are the ionization energy and the energy of the atom in
the level p, respectively, Ne is the electron number density andσion (p,ν) is the photoionization
cross section.
Equation (2) can be arranged as:
ρR p(v, T) =2 r 2 π h3 m3/2e 2p2( 1 kT) 3/2 σion(p, v)Ne2 hv3 c2 exp(− hv −(Ei−Ep) kT ) (3)
Considering the high density due to the plasma confinement in water, only a few levels are available and they are very close to the ionization energy [43]:
exp(−hv −(Ei −Ep)
kT )≈ exp(− hv
Consequently, Equation (3) becomes: ρR p(v, T) =2 q 2 π h 3 m3/2e 2p 2(1 kT) 3/2 σion(p, v)N2ehv 3 c2 exp(−kThv) = = q 2 π h 3 m3/2e 2p2(kT1)3/2σion(p, v)N2eBv(v, T) (5)
The photoionization cross section at high density conditions, shows a smooth dependence on wavelength [42], so that its contribution can be considered constant in the range of photon energy investigated in the visible range, i.e., 1–3 eV. In the frame of this assumption and since, at fixed time Ne
and T are constant, Equation (5) becomes: ρR
p(v, T) =G(T, Ne)·Bv(v, T) (6)
where G is a constant with respect to wavelength at each time of plasma lifetime since it represents a function depending on detection efficiency, on the electron number density and on plasma temperature, but not on the wavelength. Therefore, within the previous approximation, by integrating Equation (6) with dν, a Planck-like distribution can be obtained. This may justify the observation of the Planck-like distribution of plasma emission in several experiments [58–60] although the plasma is not a blackbody. Recently temporally and spatially resolved OES has been applied during LAL for the production of metallic NPs in order to estimate the temperature and density maps with data processing as described in detail in [60].
As an example, Figure3shows the temperature map of a plasma induced on an aluminum target immersed in water. By the inspection of the figure it can be observed that in the core of the plasma, where the number density is very high [60,61], there is a decrease in temperature beyond the condensation temperature of most of the metals. As a matter of fact, the high pressure of the plasma at this condition, i.e., under water confinement, allows for condensation at higher temperatures than those in standard condition. This observation indicates that NPs can be formed in the bulk of the plasma and not only in the border of the plasma where the plasma cools down fast as a consequence of the rapid exchange of energy between the plasma and the surrounding liquid. It may be assumed that also the material on the plasma border condensates in particle, but being this zone of the plasma out of equilibrium, it produces particles with various sizes and shapes. On the contrary, particles formed in the bulk of the plasma, being the result of thermodynamic equilibrium between the processes of growth and the evaporation, are characterized by a spherical shape with narrow size distribution.
Nanomaterials 2020, 10, x 6 of 50
investigated in the visible range, i.e., 1–3 eV. In the frame of this assumption and since, at fixed time
Ne and T are constant, Equation (5) becomes:
pR
(v,T ) = G(T, N
e) B
v(v,T )
(6)where G is a constant with respect to wavelength at each time of plasma lifetime since it represents a function depending on detection efficiency, on the electron number density and on plasma temperature, but not on the wavelength. Therefore, within the previous approximation, by integrating Equation (6) with dν, a Planck-like distribution can be obtained. This may justify the observation of the Planck-like distribution of plasma emission in several experiments [58–60] although the plasma is not a blackbody.
Recently temporally and spatially resolved OES has been applied during LAL for the production of metallic NPs in order to estimate the temperature and density maps with data processing as described in detail in [60].
As an example, Figure 3 shows the temperature map of a plasma induced on an aluminum target immersed in water. By the inspection of the figure it can be observed that in the core of the plasma, where the number density is very high [60,61], there is a decrease in temperature beyond the condensation temperature of most of the metals. As a matter of fact, the high pressure of the plasma at this condition, i.e., under water confinement, allows for condensation at higher temperatures than those in standard condition. This observation indicates that NPs can be formed in the bulk of the plasma and not only in the border of the plasma where the plasma cools down fast as a consequence of the rapid exchange of energy between the plasma and the surrounding liquid. It may be assumed that also the material on the plasma border condensates in particle, but being this zone of the plasma out of equilibrium, it produces particles with various sizes and shapes. On the contrary, particles formed in the bulk of the plasma, being the result of thermodynamic equilibrium between the processes of growth and the evaporation, are characterized by a spherical shape with narrow size distribution.
Figure 3. (a) Image of laser-induced plasma on Al target immersed in water, acquired with ICCD
camera. (b) Plasma temperature 2D map calculated as reported in [43] (Al target in MilliQ water, Elaser
= 270 mJ, laser crater size = 1 ± 0.2 mm, λlaser = 532 nm, laser pulse duration = 6 ns, laser frequency = 10
Hz, delay time = 40 ns, gate width = 20 ns). (c) 3D representation of plasma temperature in (b).
The mechanism of growth in the plasma phase during ns-LAL has been investigated with a theoretical model in [61] where the competition between thermodynamic condensation and electrostatic growth has been investigated. The result shows that at the electron number density usually found in ns-LIP in water, the charging of seeds and particles occurs at ps-time scale. This observation suggests that as soon as small clusters are formed because of the high-density gradient at the initial stage of laser ablation, electrons attach. Consequently, the particles become negatively charged and attract ions in the plasma. Electron charging and ion implantation allows the particles to grow until the rate of growth is balanced by the evaporation process due to the high temperature in the plasma (4000–6000 K). In this scenario, the equilibrium between electrostatic growth and
Figure 3. (a) Image of laser-induced plasma on Al target immersed in water, acquired with ICCD camera. (b) Plasma temperature 2D map calculated as reported in [43] (Al target in MilliQ water, Elaser= 270 mJ, laser crater size = 1 ± 0.2 mm, λlaser= 532 nm, laser pulse duration = 6 ns, laser frequency= 10 Hz, delay time = 40 ns, gate width = 20 ns). (c) 3D representation of plasma temperature in (b).
Nanomaterials 2020, 10, 2317 7 of 50
The mechanism of growth in the plasma phase during ns-LAL has been investigated with a theoretical model in [61] where the competition between thermodynamic condensation and electrostatic growth has been investigated. The result shows that at the electron number density usually found in ns-LIP in water, the charging of seeds and particles occurs at ps-time scale. This observation suggests that as soon as small clusters are formed because of the high-density gradient at the initial stage of laser ablation, electrons attach. Consequently, the particles become negatively charged and attract ions in the plasma. Electron charging and ion implantation allows the particles to grow until the rate of growth is balanced by the evaporation process due to the high temperature in the plasma (4000–6000 K). In this scenario, the equilibrium between electrostatic growth and thermodynamic evaporation set the NPs dimensions and induces the typical spherical shape observed after the LAL process.
Recently it has been shown, that the charging effect of the plasma is involved also in the subsequent steps of LAL process [62,63]. As a matter of fact, when the particle exits the plasma, it is still with an excess of electrons. This charge, in the order of a few nC for the entire NPs production per laser shot, allows the repulsion between the particles and preserves the NPs from massive aggregation.
The cavitation bubble: The fast transfer of energy from the plasma to the surrounding water induces the formation of a thin layer of vapor around the plasma border with high temperature and high pressure. In order to reach the equilibrium with the liquid, the vapor expands, producing a cavitation bubble. In the upper part of Figure4, an example of the temporal evolution of a laser-induced bubble during silver ablation is reported. The bubble expands until it reaches the equilibrium with the surrounding liquid. Due to the fast expansion, the liquid at the border of the cavitation is compressed and when the pressure of the bubble reaches the minimum at the maximum of the expansion, the bubble is compressed and its shrinking stage begins. During this stage, the bubble increases its pressure again and impacts the target, eventually pushing back the material to the target surface. There are several models describing the evolution of a laser-induced bubble [64,65]. Most of these descriptions are based on a theoretical model simulating the bubble radius evolution, assuming the mass and momentum conservation equations in the liquid phase, and that the vapor pressure inside the bubble is balanced by the pressure on the liquid side of bubble wall [66]. As an example, neglecting the evaporation/condensation effects as the time associated to bubble dynamics is shorter than that of these processes, the Keller–Miksis formulation can be applied with the following equations [66]:
(1 − M)ρlRR..+3 2(1 − M 3)ρl . R2= (1+M+R cl d dt)(pg(t)− 2σl R − 4ηl . R R−p∞) (7)
where R is the radius of the bubble and each dot indicates a time derivative,ρlis the liquid water density,
M is the bubble-wall Mach number, clis the sound speed in the liquid, pgand p∞are, respectively,
the gas pressure inside the bubble and the background static pressure andσlandηlare, respectively,
the surface tension and dynamic viscosity of liquid water. The equation can be resolved considering the weakly compressible equation of state for the liquid density, and assuming an adiabatic van der Waals equation of state for a spherical bubble so that the vapor pressure and temperature are related to the radius by the following equations:
pg(t) = (p∞+2σl R )( R3∞−h3 R3−h3) γ (8) T(t) =T∞( R3∞−h3 R3−h3) γ−1 (9) where R∞ is the radius at which the pressure inside the bubble corresponds to p∞, h is the radius
determined by the excluded volume of the water molecules andγ = Cp/Cv= (2 + N)/N is the ratio of the
specific heats, with N the number of degrees of freedom. The model starts assuming the initial radius equal to that of the laser spot and the gas inside the bubble in thermal equilibrium with the liquid.
Although this simulation does not take into account the interaction of laser-induced plasma with the vapor inside the bubble, it provides an acceptable qualitative description of bubble evolution.
Nanomaterials 2020, 10, x 8 of 50
Figure 4. Time resolved shadowgraphy images of the laser induced bubble on Ag target immersed in
water (a) and the corresponding bubble radius compared with the number of ejected Ag NPs during the bubble evolution (as calculated in [44]) (b). Experimental conditions: laser energy = 108 mJ, laser frequency = 2 Hz, laser crater size = 1 ± 0.2 mm, ICCD gate width = 20 µs and 5 accumulations for each image.
As a result of the charging effect on the NPs produced in the plasma phase due to the electrostatic
repulsion, NP clouds expand as well. If the NPs store enough charge immediately after the plasma
stage, the induced electrostatic pressure can be higher than the pressure at the water/vapor boundary
and NPs exit from the cavitation bubble and are released in solution. Recently, this phenomenon has
been investigated in detail in [62] and it has been found that the two main streams of NPs ejection
due to the electrostatic repulsion within the particle cloud in the bubble occurs immediately after the
beginning of the expansion phase and, in a greater extent, between the collapsing stage and the
bubble rebound [62]. Figure 4 shows the number of ejected AgNPs during the cavitation bubble
evolution compared with the measured bubble radius, using Equations (8) and (9). The number of
the ejected AgNPs has been calculated as reported in [62].
It would be important to underline that the spherical NPs with a size around 10–20 nm are not
the only particles produced. Indeed, in terms of mass, the fraction of material converted in this kind
of NPs is a minor component. Larger particles are produced at the border of the plasma during the
energy exchange from plasma to the surrounding liquid as mentioned below. Other bigger NPs can
be the result of aggregation of the spherical NPs during the collapse stage of the bubble. The NPs that
are not ejected because of the electrostatic repulsion can remain trapped in the bubble, therefore they
can be then compressed, because of the high pressure of the collapse stage, rearranging their structure
in aggregates composed by individual spherical NPs. Finally, as an intrinsic result of the laser
ablation, debris of several hundred of nm with the same morphology of the irradiated target can be
found on the bottom of the ablation cell.
All the discussion here refers to ns laser ablation and it gives a simplified scenario of the
processes involved in the production of NPs of high quality in terms of size distribution and chemical
purity of the colloidal solution. In order to increase the production rate, in recent years, different
approaches have been developed based on K-Hz laser source, ultrashort laser pulse and target shape,
as will be discussed in Section 2.2.
Figure 4.Time resolved shadowgraphy images of the laser induced bubble on Ag target immersed in water (a) and the corresponding bubble radius compared with the number of ejected Ag NPs during the bubble evolution (as calculated in [44]) (b). Experimental conditions: laser energy= 108 mJ, laser frequency= 2 Hz, laser crater size = 1 ± 0.2 mm, ICCD gate width = 20 µs and 5 accumulations for each image.
As a result of the charging effect on the NPs produced in the plasma phase due to the electrostatic repulsion, NP clouds expand as well. If the NPs store enough charge immediately after the plasma stage, the induced electrostatic pressure can be higher than the pressure at the water/vapor boundary and NPs exit from the cavitation bubble and are released in solution. Recently, this phenomenon has been investigated in detail in [62] and it has been found that the two main streams of NPs ejection due to the electrostatic repulsion within the particle cloud in the bubble occurs immediately after the beginning of the expansion phase and, in a greater extent, between the collapsing stage and the bubble rebound [62]. Figure4shows the number of ejected AgNPs during the cavitation bubble evolution compared with the measured bubble radius, using Equations (8) and (9). The number of the ejected AgNPs has been calculated as reported in [62].
It would be important to underline that the spherical NPs with a size around 10–20 nm are not the only particles produced. Indeed, in terms of mass, the fraction of material converted in this kind of NPs is a minor component. Larger particles are produced at the border of the plasma during the energy exchange from plasma to the surrounding liquid as mentioned below. Other bigger NPs can be the result of aggregation of the spherical NPs during the collapse stage of the bubble. The NPs that are not ejected because of the electrostatic repulsion can remain trapped in the bubble, therefore they can be then compressed, because of the high pressure of the collapse stage, rearranging their structure in aggregates composed by individual spherical NPs. Finally, as an intrinsic result of the laser ablation,
Nanomaterials 2020, 10, 2317 9 of 50
debris of several hundred of nm with the same morphology of the irradiated target can be found on the bottom of the ablation cell.
All the discussion here refers to ns laser ablation and it gives a simplified scenario of the processes involved in the production of NPs of high quality in terms of size distribution and chemical purity of the colloidal solution. In order to increase the production rate, in recent years, different approaches have been developed based on K-Hz laser source, ultrashort laser pulse and target shape, as will be discussed in Section2.2.
2.2. Upscaling of Laser Synthesis of Colloids
As can be derived from the previous section, understanding the mechanisms of a process enables its control, which can then lead to a precise experimental design. This section reviews process parameters of LAL that are of great importance for the scale-up of this process [66]. Additionally, cross-dependencies between the single parameters are pointed out.
2.2.1. Scaling and Control Factors for Laser Ablation in Liquids
LAL process parameters are initial, adjustable, and related to components of the process, which are the laser, the components of the optical setup, the ablation chamber, the fluid flow elements, the ablation target, and the liquid environment. The process result of interest is the NP productivity, which can be defined as volume of product achieved in a defined irradiation time (considering energy efficiency, additional reference values like the laser power become relevant as well). In literature, the ablation rate is frequently used for productivity comparisons. It corresponds to the loss of target volume achieved in a defined irradiation time. By assuming a 100% conversion of ablated matter to NPs, the definitions of productivity and ablation rate are identical.
A process starts with the input of initial parameters and creates an output, which comprises the results. In between those process parameters and results, variables can be defined. They generally depend on multiple process parameters of different components of the setup and bundle their impact on the process. The dependency graph shown in Figure5displays the compilation of the process parameters of LAL clustered and assigned to the different components of the LAL setup. In addition, the cross-dependencies of parameters, found in the process variables, and finally the process result(s) are also illustrated in the figure.
Nanomaterials 2020, 10, x 9 of 50
2.2. Upscaling of Laser Synthesis of Colloids
As can be derived from the previous section, understanding the mechanisms of a process enables its control, which can then lead to a precise experimental design. This section reviews process parameters of LAL that are of great importance for the scale-up of this process [66]. Additionally, cross-dependencies between the single parameters are pointed out.
2.2.1. Scaling and Control Factors for Laser Ablation in Liquids
LAL process parameters are initial, adjustable, and related to components of the process, which are the laser, the components of the optical setup, the ablation chamber, the fluid flow elements, the ablation target, and the liquid environment. The process result of interest is the NP productivity, which can be defined as volume of product achieved in a defined irradiation time (considering energy efficiency, additional reference values like the laser power become relevant as well). In literature, the ablation rate is frequently used for productivity comparisons. It corresponds to the loss of target volume achieved in a defined irradiation time. By assuming a 100% conversion of ablated matter to NPs, the definitions of productivity and ablation rate are identical.
A process starts with the input of initial parameters and creates an output, which comprises the results. In between those process parameters and results, variables can be defined. They generally depend on multiple process parameters of different components of the setup and bundle their impact on the process. The dependency graph shown in Figure 5 displays the compilation of the process parameters of LAL clustered and assigned to the different components of the LAL setup. In addition, the cross-dependencies of parameters, found in the process variables, and finally the process result(s) are also illustrated in the figure.
The use of different colors supports the readability of the dependency graph. Each time two or more parameters or variables interdepend, their colors add up. For example, the input parameter pulse duration of the cluster laser influences directly the output productivity but also indirectly the variable intensity. The colors of pulse duration (green) and energy density (wine) add up to dark brown, which represents the variable intensity. In the following sections, the impact of single input parameters and also the assembled variables on the NP productivity are evaluated. Note that product properties such as the particle size distribution also represent an output of the process but are not considered in detail here.
Figure 5. Dependence graph of process parameters, arranged in component clusters, relevant to LAL
up-scaling. If not marked differently by an arrow, the flow goes from the left to the right in horizontal direction and from the outside to the inside in vertical direction. Colors were used only for an improved overview. Reprinted with permission from [66].
Laser Wavelength: The laser wavelength essentially influences the ablation rate in LAL [54,67– 70]. All studies agree independently from the ablated matter, the chosen liquid and the used laser
Figure 5.Dependence graph of process parameters, arranged in component clusters, relevant to LAL up-scaling. If not marked differently by an arrow, the flow goes from the left to the right in horizontal direction and from the outside to the inside in vertical direction. Colors were used only for an improved overview. Reprinted with permission from [66].
The use of different colors supports the readability of the dependency graph. Each time two or more parameters or variables interdepend, their colors add up. For example, the input parameter pulse duration of the cluster laser influences directly the output productivity but also indirectly the variable intensity. The colors of pulse duration (green) and energy density (wine) add up to dark brown, which represents the variable intensity. In the following sections, the impact of single input parameters and also the assembled variables on the NP productivity are evaluated. Note that product properties such as the particle size distribution also represent an output of the process but are not considered in detail here.
Laser Wavelength: The laser wavelength essentially influences the ablation rate in LAL [54,67–70]. All studies agree independently from the ablated matter, the chosen liquid and the used laser pulse duration on a higher productivity by using IR laser light compared to UV or Vis. The laser fluence was kept at identical or at least comparable values in all studies. However, LAL was applied as a batch approach in all investigations, and NP-induced shielding due to Rayleigh scattering [71] most likely caused the differences in productivity. In addition, a higher absorption cross-section is given for NPs of nearly all metals at UV or Vis wavelengths compared to IR wavelengths [72]. This is strongly indicated by the observed smaller NP diameters at shorter applied wavelengths [54,69,70], which may be the result of fragmentation induced by subsequent laser pulses during batch-LAL.
Schwenke et al. [54] determined the product concentration during picosecond-pulsed LAL at different process times for IR and Vis wavelengths and three different metals. They found that the concentration was comparable for both wavelengths at short process times. However, it increased differently with time or the NP concentration, and less strongly at short wavelengths in all cases. These results clearly demonstrate the shielding effect of the colloidal NPs.
Semerok et al. [73] found higher ablation rates by using shorter wavelengths in the picosecond and nanosecond-pulsed laser ablation in air for several metals. The ablation rate increased stronger for most metals when switching from the Vis to the UV range then from the IR to the Vis range, which was in good agreement with the interband absorption of the targets. Similar results were also found by Stafe et al. [74]. Naturally, the individual light absorption properties of bulk materials should influence the number of absorbed photons of different wavelengths and provide optimums for high ablation rates. However, the optical properties of an irradiated surface significantly depend on its temperature [75], roughness [76,77] and oxidation state/degree [78]. For example, Letzel et al. [79] recently observed a significant increase in the productivity of the laser ablation of silver in water during the first 250 laser pulses irradiating an initially smooth target surface. Temperature, roughness and oxidation state/degree of the target surface change differently strong during LAL and depend strongly on the ablated material. This limits the a priori choice of the most productive wavelength for LAL according to the standard light absorption spectrum of the material to be ablated.
It may be concluded that higher ablation rates for metals generally occur at UV laser wavelengths due to the interband absorption. However, the productivity at Vis and IR wavelengths can be comparable. A strong dependence on the metal, the laser fluence and also the laser pulse duration exists. In case of high colloidal concentrations, the higher absorbance of short wavelengths by the particles can make LAL of metals at short wavelengths less productive. These relations are summarized in Figure6. Note that for semiconducting and dielectric materials the initial ablation rate in absence of a light-extinction colloid can be higher at red and near IR wavelengths under specific conditions due to differences in the ablation mechanism compared to metals [80,81].
Laser Fluence: The laser fluence or energy density describes the pulse energy penetrating the area of the effective laser spot on the surface of the ablation target. A variation of the laser fluence may be realized by either varying the pulse energy [67,70,71,82–90] or the spot area via the working distance [67,85,91]. Different results by both procedures can occur due to variations of the beam propagation within the liquid phase. Furthermore, the distribution of the laser energy density in the spot of a Gaussian beam differs. Naturally, the ablation rate in LAL follows a logarithmic growth when increasing the laser fluence of femtosecond [90], picosecond [82,88] and nanosecond [85] pulses.
Nanomaterials 2020, 10, 2317 11 of 50
At very low values of the laser fluence, a less productive ablation regime prevails [82], which is not considered here. The logarithmic dependence is also known for laser ablation in a gaseous environment [91–95].
Nanomaterials 2020, 10, x 10 of 50
pulse duration on a higher productivity by using IR laser light compared to UV or Vis. The laser
fluence was kept at identical or at least comparable values in all studies. However, LAL was applied
as a batch approach in all investigations, and NP-induced shielding due to Rayleigh scattering [71]
most likely caused the differences in productivity. In addition, a higher absorption cross-section is
given for NPs of nearly all metals at UV or Vis wavelengths compared to IR wavelengths [72]. This
is strongly indicated by the observed smaller NP diameters at shorter applied wavelengths [54,69,70],
which may be the result of fragmentation induced by subsequent laser pulses during batch-LAL.
Schwenke et al. [54] determined the product concentration during picosecond-pulsed LAL at
different process times for IR and Vis wavelengths and three different metals. They found that the
concentration was comparable for both wavelengths at short process times. However, it increased
differently with time or the NP concentration, and less strongly at short wavelengths in all cases.
These results clearly demonstrate the shielding effect of the colloidal NPs.
Semerok et al. [73] found higher ablation rates by using shorter wavelengths in the picosecond
and nanosecond-pulsed laser ablation in air for several metals. The ablation rate increased stronger
for most metals when switching from the Vis to the UV range then from the IR to the Vis range, which
was in good agreement with the interband absorption of the targets. Similar results were also found
by Stafe et al. [74]. Naturally, the individual light absorption properties of bulk materials should
influence the number of absorbed photons of different wavelengths and provide optimums for high
ablation rates. However, the optical properties of an irradiated surface significantly depend on its
temperature [75], roughness [76,77] and oxidation state/degree [78]. For example, Letzel et al. [79]
recently observed a significant increase in the productivity of the laser ablation of silver in water
during the first 250 laser pulses irradiating an initially smooth target surface. Temperature, roughness
and oxidation state/degree of the target surface change differently strong during LAL and depend
strongly on the ablated material. This limits the a priori choice of the most productive wavelength for
LAL according to the standard light absorption spectrum of the material to be ablated.
It may be concluded that higher ablation rates for metals generally occur at UV laser
wavelengths due to the interband absorption. However, the productivity at Vis and IR wavelengths
can be comparable. A strong dependence on the metal, the laser fluence and also the laser pulse
duration exists. In case of high colloidal concentrations, the higher absorbance of short wavelengths
by the particles can make LAL of metals at short wavelengths less productive. These relations are
summarized in Figure 6. Note that for semiconducting and dielectric materials the initial ablation
rate in absence of a light-extinction colloid can be higher at red and near IR wavelengths under
specific conditions due to differences in the ablation mechanism compared to metals [80,81].
Figure 6. Schematic of the laser ablation of metals in liquids at different wavelengths in absence (a)
and presence (b) of colloidal particles. Equal energy densities were assumed for all laser spots. The
diagram in (c) (linear scales for both axes) demonstrates the effect of an increasing colloidal
concentration on the NP productivity in LAL at different wavelengths. The trends shown in the
diagram represent the interpretation of the combined results of the discussed literature in Section
2.2.1. Reprinted with permission from [66].
Laser Fluence: The laser fluence or energy density describes the pulse energy penetrating the
area of the effective laser spot on the surface of the ablation target. A variation of the laser fluence
may be realized by either varying the pulse energy [67,70,71,82
–
90] or the spot area via the working
distance [67,85,91]. Different results by both procedures can occur due to variations of the beam
Figure 6.Schematic of the laser ablation of metals in liquids at different wavelengths in absence (a) and
presence (b) of colloidal particles. Equal energy densities were assumed for all laser spots. The diagram in (c) (linear scales for both axes) demonstrates the effect of an increasing colloidal concentration on the NP productivity in LAL at different wavelengths. The trends shown in the diagram represent the interpretation of the combined results of the discussed literature in Section2.2.1. Reprinted with permission from [66].
The onset of the saturation occurs at lower laser fluences for shorter laser pulse durations [90], which may be due to higher laser intensities and the related optical breakdown of the liquid. The lowered thermal impact on the material, and therefore stronger independence of laser penetration depths and fluence at shorter laser pulse length, may also play a role here. In addition, the type of the ablated material influences the onset of the saturation in the ablation rate [83,85,92] by different laser fluence thresholds and penetration depths. For example, Hahn et al. [83] could not increase the productivity of the laser ablation of silver in water in the pulse energy range of 100 to 500 µJ but nearly tripled the productivity of the ablation of cobalt by increasing the pulse energy from 100 to 200 µJ before reaching saturation. The working distance was kept constant by the authors.
The conventional definition of productivity in LAL (volume ablation per time) used in this manuscript can also be related to the applied laser energy. This extended definition allows an assessment of the energy efficiency, which seems reasonable from an economic and ecological point of view. Neuenschwander et al. [96] demonstrated theoretically and experimentally that a maximum energy efficiency of ultrashort-pulsed laser ablation is reached at an optimum fluence, which is the material-specific threshold fluence of laser ablation multiplied by e2. While the authors of the previously mentioned study focused on laser ablation in air, Streubel et al. [44] also verified the correlation for laser ablation in liquids.
In conclusion, the volume ablation rate logarithmically scales with the laser fluence. The progress of the logarithmic function further depends on the ablation mechanism, the surrounding liquid, and the target material. If productivity in LAL is further related to the invested laser pulse energy, an optimum fluence can be found. Figure7illustrates the typical dependency of the NP productivity on the laser fluence.
Laser Pulse Duration: The mechanism of laser ablation crucially depends on the pulse duration. The influence of moderate fluences on the ablation rate of metals in gaseous atmosphere can be well described by the two-temperature model [97] for femtosecond and picosecond pulses and by a simple evaporation model at longer durations [95]. Extensions and refinements of the two-temperature model even allow accurate predictions of the ablation rate in the femtosecond-pulsed laser ablation of dielectrics [98] or of the low-fluence regime in the picosecond-pulsed laser ablation of metals [99]. Exchanging the gaseous ambient by a liquid causes a more intense cooling of the ablation target and
Nanomaterials 2020, 10, 2317 12 of 50
of the laser-induced plasma, which may affect the ablation process but should neither change the principal ablation mechanism nor invalidate the describing models.
propagation within the liquid phase. Furthermore, the distribution of the laser energy density in the
spot of a Gaussian beam differs. Naturally, the ablation rate in LAL follows a logarithmic growth
when increasing the laser fluence of femtosecond [90], picosecond [82,88] and nanosecond [85] pulses.
At very low values of the laser fluence, a less productive ablation regime prevails [82], which is not
considered here. The logarithmic dependence is also known for laser ablation in a gaseous
environment [91–95].
The onset of the saturation occurs at lower laser fluences for shorter laser pulse durations [90],
which may be due to higher laser intensities and the related optical breakdown of the liquid. The
lowered thermal impact on the material, and therefore stronger independence of laser penetration
depths and fluence at shorter laser pulse length, may also play a role here. In addition, the type of
the ablated material influences the onset of the saturation in the ablation rate [83,85,92] by different
laser fluence thresholds and penetration depths. For example, Hahn et al. [83] could not increase the
productivity of the laser ablation of silver in water in the pulse energy range of 100 to 500 µJ but
nearly tripled the productivity of the ablation of cobalt by increasing the pulse energy from 100 to
200 µJ before reaching saturation. The working distance was kept constant by the authors.
The conventional definition of productivity in LAL (volume ablation per time) used in this
manuscript can also be related to the applied laser energy. This extended definition allows an
assessment of the energy efficiency, which seems reasonable from an economic and ecological point
of view. Neuenschwander et al. [96] demonstrated theoretically and experimentally that a maximum
energy efficiency of ultrashort-pulsed laser ablation is reached at an optimum fluence, which is the
material-specific threshold fluence of laser ablation multiplied by e
2. While the authors of the
previously mentioned study focused on laser ablation in air, Streubel et al. [44] also verified the
correlation for laser ablation in liquids.
In conclusion, the volume ablation rate logarithmically scales with the laser fluence. The
progress of the logarithmic function further depends on the ablation mechanism, the surrounding
liquid, and the target material. If productivity in LAL is further related to the invested laser pulse
energy, an optimum fluence can be found. Figure 7 illustrates the typical dependency of the NP
productivity on the laser fluence.
Figure 7. Illustration of the laser ablation of metals in liquids at different fluences varied by the pulse
energy (a) and the working distance (b). The diagram in (c) demonstrates how the fluence variation
caused by different initial spot sizes affects the NP productivity in LAL in the high-fluence regime. A
Gaussian beam profile was assumed. In (c), both axes scale linearly. The trends shown in the diagram
represent the interpretation of the combined results of the discussed literature in Section 2.2.1. The
arrow and the less dense (dashed) line style in (c) indicate the trend in productivity induced by a
smaller initial spot size. Reprinted with permission from [66].
Laser Pulse Duration: The mechanism of laser ablation crucially depends on the pulse duration.
The influence of moderate fluences on the ablation rate of metals in gaseous atmosphere can be well
described by the two-temperature model [97] for femtosecond and picosecond pulses and by a simple
evaporation model at longer durations [95]. Extensions and refinements of the two-temperature
model even allow accurate predictions of the ablation rate in the femtosecond-pulsed laser ablation
of dielectrics [98] or of the low-fluence regime in the picosecond-pulsed laser ablation of metals [99].
Exchanging the gaseous ambient by a liquid causes a more intense cooling of the ablation target and
of the laser-induced plasma, which may affect the ablation process but should neither change the
principal ablation mechanism nor invalidate the describing models.
Figure 7.Illustration of the laser ablation of metals in liquids at different fluences varied by the pulse
energy (a) and the working distance (b). The diagram in (c) demonstrates how the fluence variation caused by different initial spot sizes affects the NP productivity in LAL in the high-fluence regime. A Gaussian beam profile was assumed. In (c), both axes scale linearly. The trends shown in the diagram represent the interpretation of the combined results of the discussed literature in Section2.2.1. The arrow and the less dense (dashed) line style in (c) indicate the trend in productivity induced by a smaller initial spot size. Reprinted with permission from [66].
Conclusively, a strong impact of the pulse duration on the ablation rate exists. A fair experimental comparison of the ablation rate with laser pulses with durations ranging from some 10 femtosecond to some 100 ns is not accessible due to the lack of suitable laser technology. However, if a wide-range variation of the pulse duration would be possible (keeping the pulse energy constant), a more effective use of the energy of single laser pulses should occur at shorter pulse durations due to the reduced heat loss. However, the optical breakdown of the liquid in front of the ablation target as well as non-linear optical effects such as the optical Kerr leading to filamentation limits the applicable fluence stronger at shorter laser pulses due to their higher intensities.
Riabinina et al. [90] investigated the NP productivity of the LAL of gold in an aqueous sodium citrate solution at different pulse durations between 40 fs and 200 ps but a constant fluence of 50 mJ/cm2.
An optimal productivity was found at a pulse duration of 2 ps. The authors expected a decrease in the productivity at lower pulse durations to be caused by the optical breakdown of the liquid. A loss of productivity at higher pulse durations was expected to originate from heat losses in the target and laser-induced shielding effects. During their experiments, the laser pulse repetition rate was kept constant and the scanning speed was kept at zero, which makes shielding effects at longer pulse durations realistic. Interestingly, Sakka et al. [100] also observed a decreasing ablation rate at an increasing pulse duration during the nanosecond-pulsed laser ablation of copper in water, although this effect did not occur during laser ablation in air. This finding confirms the origin of a negative impact of an increasing pulse duration on the ablation rate in the presence of the liquid environment. In another study, Dittrich et al. compared the absolute and power-specific productivity of LAL of gold in water for a compact, a middle class and a high-end laser system [101]. The authors applied LAL in a liquid flow and with sufficient laser scanning to limit shielding effects. They found that even though the absolute productivity was the lowest for the low-power, compact laser system, power-specific productivity was much higher compared to the other systems. All three laser systems differed in the laser pulse duration, which was 1 ns for the compact, 5 ns for the middle-class and 3 ps for the high-end laser system, respectively.
Further details on the effect of the pulse duration on the ablation rate can be derived from studies on the laser ablation of metals in air and vacuum. Schille et al. [102] found a slight decrease in the ablation rate of copper during an increase of the pulse duration from 200 fs to 4 ps. A further increase of the pulse duration up to 10 ps significantly reduced the ablation rate. As demonstrated by Jaeggi et al. [103], the threshold fluence for ablation increases and the penetration depth decreases
Nanomaterials 2020, 10, 2317 13 of 50
in the ablation of copper and steel in air when increasing the pulse duration from 10 ps to 50 ps. Naturally, both effects negatively affect the ablation rate. If the pulse duration is further increased from the picosecond to the few nanosecond time regime, the threshold fluence for ablation becomes constant [104]. In addition, the ablation rate becomes independent from the optical penetration depths at these longer pulse durations due to the more thermal ablation mechanism [94]. Minor changes in the ablation rate can be assumed in the transition regime from picosecond to nanosecond pulse durations. However, the thermal nature of the laser ablation intensifies at even longer pulse durations. Heat losses get more dominant and the ablation rate is consequently reduced [105]. In addition, the lifetime of the laser-induced plasma and the pulse duration act on a comparable time scale for laser pulses longer than some ten or hundred nanoseconds, which intensifies a self-induced plasma shielding of single laser pulses.
In conclusion, an optimum in the pulse duration for achieving a maximized ablation rate in LAL exists. It is probably located in the order of single picoseconds [86]. As the productivity limitation at lower pulse durations stems from the optical breakdown of the liquid, the optimum of the pulse duration strongly depends on the laser intensity, wavelength and the liquid. Figure8illustrates the thermal loss and the pulse self-shielding by its induced plasma in dependence on the pulse duration regime. Furthermore, the dependence of the productivity on the pulse duration is shown qualitatively by excluding the optical breakdown of the liquid. The optical breakdown finds consideration in the following section on the laser intensity.
Nanomaterials 2020, 10, x 13 of 50
Figure 8. Illustration of the laser ablation of metals in liquids at different pulse durations neglecting
non-linear effects. Thermal losses (a) and self-shielding by plasma (b) of pulses of different duration
regimes. The diagram in (c) demonstrates how the coupling constant of electrons and phonons affects
the NP productivity in LAL, depending on the laser pulse duration. A change of the coupling constant
was assumed to have no influence on other material properties. In (c), the y-axis scales linearly. The
trends shown in the diagram represent the interpretation of the combined results of the discussed
literature in Section 2.2.1. In (c), the arrow and the less dense (dashed) line style indicate the effect of
a longer coupling time of electrons and phonons of the target material on productivity. Reprinted
with permission from [66].
Laser Intensity: Since there is a tight connection between laser intensity, laser fluence and pulse
duration, a differentiation between each contribution is difficult. The optical breakdown of the liquid
environment, which may significantly reduce the ablation rate, is usually related to the laser intensity.
Additionally, self-focusing of the laser beam may be induced by the optical Kerr effect and
consequently increase the laser intensity. This may promote filamentation of the laser beam or an
optical breakdown of the liquid. However, the intensity thresholds needed for the optical breakdown
of water and the filament formation are comparable for femtosecond-pulsed laser radiation [106]. For
picosecond pulses, the intensity threshold for filamentation is generally much higher than for the
optical breakdown of water and becomes comparable only at low focusing angles [107]. Thus,
filamentation plays no or only a minor role in LAL.
As described by Noack and Vogel [108], the laser-induced cascade of ionization must reach a
specific density of free electrons to provoke the breakdown of the liquid in dependence on the pulse
duration and wavelength. It is thus a competition of the recombination and ionization rate of free
electrons, ions and atoms. Due to higher electron densities, plasmas induced by nanosecond laser
pulses more effectively shield later irradiation than plasmas induced by picosecond pulses or long
femtosecond pulses [109]. Additionally, one to two magnitudes of higher intensity thresholds for the
optical breakdown of water are valid for femtosecond (10
12W/cm
2) compared to nanosecond (10
10W/cm
2) pulses.
The intensity threshold for the optical breakdown of the liquid environment further depends on
the chemical nature and the purity of the liquid. In the case of linear hydrocarbons, Fujii et al. [110]
found a decrease in the breakdown threshold for an increasing chain length or molecular weight
using IR, nanosecond pulses. This is probably due to the increasing absorption cross-section of the
hydrocarbons. The threshold intensities for the optical breakdown of different hydrocarbon
aromatics at IR, nanosecond pulses are in the order of 10
12W/cm
2[111]. Kovalchuk et al. [112]
demonstrated, temporally and spatially resolved, how particulate impurities in tap water and
different alcohols act as seeds for the plasma formation in the laser-induced optical breakdown of the
liquids. In addition, the presence of soluble species also significantly reduces the intensity threshold
of a liquid for its breakdown [113,114]. It must be assumed that NPs act as solutes or impurities in
this context. In addition, breakdown seeds may also be introduced into liquids by complexation of
solvent molecules with dissolved gases [115].
In conclusion, the optical breakdown of the liquid environment in absence of an ablation target
must be excluded for the applied laser parameters in LAL. Thereby it needs to be considered that
colloidal particles and dissolved species reduce the breakdown threshold. Whether an optical
breakdown of the liquid occurs during LAL should therefore always be checked experimentally.
Figure 8.Illustration of the laser ablation of metals in liquids at different pulse durations neglecting
non-linear effects. Thermal losses (a) and self-shielding by plasma (b) of pulses of different duration regimes. The diagram in (c) demonstrates how the coupling constant of electrons and phonons affects the NP productivity in LAL, depending on the laser pulse duration. A change of the coupling constant was assumed to have no influence on other material properties. In (c), the y-axis scales linearly. The trends shown in the diagram represent the interpretation of the combined results of the discussed literature in Section2.2.1. In (c), the arrow and the less dense (dashed) line style indicate the effect of a longer coupling time of electrons and phonons of the target material on productivity. Reprinted with permission from [66].
Laser Intensity: Since there is a tight connection between laser intensity, laser fluence and pulse duration, a differentiation between each contribution is difficult. The optical breakdown of the liquid environment, which may significantly reduce the ablation rate, is usually related to the laser intensity. Additionally, self-focusing of the laser beam may be induced by the optical Kerr effect and consequently increase the laser intensity. This may promote filamentation of the laser beam or an optical breakdown of the liquid. However, the intensity thresholds needed for the optical breakdown of water and the filament formation are comparable for femtosecond-pulsed laser radiation [106]. For picosecond pulses, the intensity threshold for filamentation is generally much higher than for the optical breakdown of water and becomes comparable only at low focusing angles [107]. Thus, filamentation plays no or only a minor role in LAL.
As described by Noack and Vogel [108], the laser-induced cascade of ionization must reach a specific density of free electrons to provoke the breakdown of the liquid in dependence on the pulse
